DE: Direct effect estimates and asymptotic variance.

In gpapadog/Interference: Causal inference with interference for realistic regimes

Direct effect estimates and asymptotic variance.

Description

Direct effect estimates and asymptotic variance.

Usage

DE(ypop, ypop_var, boots = NULL, alpha = NULL, alpha_level = 0.05)

Arguments

ypop	A matrix with rows corresponding to the potential outcome under control and treatment, and columns corresponding to the cluster-average propensity of treatment.
ypop_var	An 3-dimensional array, where the first two dimensions are equal to 2 and include the variance covariance matrix of the population average potential outcome for each alpha. Dimension 3 is alpha.
boots	The results of BootVar() function including estimates of the potential outcomes from the bootstrap samples.
alpha	The values of alpha we consider. If ypop has column names, alpha can be left null.
alpha_level	Numeric. The alpha level of the confidence intervals based on the quantiles of the bootstrap estimates.

Value

A matrix with rows including the estimate and variance of the direct effect and columns corresponding to alpha.

gpapadog/Interference documentation built on Oct. 30, 2022, 9:20 p.m.